Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Robust Cross-Etiology and Speaker-Independent Dysarthric Speech Recognition

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that a single Whisper model fine-tuned on Parkinson's dysarthric speech reaches near-typical error rates on unseen speakers and transfers without retraining to cerebral palsy and ALS speech.

desk verdict Whisper baseline on SAP-1005 is useful, but validation and test share the same speakers, undercutting the speaker-independence claim. read the letter →

arxiv 2501.14994 v1 pith:U7GY7RI4 submitted 2025-01-25 cs.SD cs.AIcs.LGeess.AS

Satwinder Singh , Qianli Wang , Zihan Zhong , Clarion Mendes , Mark Hasegawa-Johnson , Waleed Abdulla , Seyed Reza Shahamiri This is my paper · ORCID
classification cs.SDcs.AIcs.LGeess.AS
keywords dysarthricspeechrecognitionspeaker-independentASRWhisperfine-tuningAccessibilityProjectSAP-1005TORGOdatasetcross-etiologytransferParkinson'sdisease
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to show that a single speaker-independent automatic speech recognition system can handle dysarthric speech well enough to be useful, across both unseen speakers and different underlying neurological conditions. It fine-tunes a medium-sized multilingual Whisper model on the SAP-1005 corpus of Parkinson's disease speech and reports 6.99% character error rate and 10.71% word error rate on the dev-unshared set, a 60.24% relative WER improvement over the earlier wav2vec 2.0 baseline. The same model, without any retraining, reaches 25.08% CER and 39.56% WER on the TORGO corpus of cerebral palsy and ALS speech. If the data split is truly speaker- and prompt-disjoint, this establishes a strong baseline for speaker-independent dysarthric ASR and suggests that acoustic regularities are shared across etiologies.

What carries the argument

The workhorse is the Whisper encoder-decoder transformer, in its medium multilingual variant, fine-tuned on SAP-1005 and decoded with beam search (num_beams=10, no_repeat_ngram_size=3, length_penalty=1.0). Because Whisper's receptive field is 30 seconds, utterances longer than 30 seconds are split into 30-second chunks with a 5-second overlap and decoded separately, then the transcripts are concatenated; this chunking is what keeps long spontaneous prompts from being lost to repetition hallucinations. The fine-tuned model is then applied unchanged to TORGO continuous speech.

What would settle it

Re-identify every speaker ID and text prompt in the SAP-1005 dev-unshared set and check whether any appears in the training partition; then re-run the same Whisper fine-tune on a verified disjoint split and compare the CER and WER to 6.99% and 10.71%.

Watch

Extended reading notes

Core claim

The central claim is that one fine-tuned Whisper model generalizes across dysarthric speakers in a way that conventional speaker-dependent systems do not. Trained only on PD speech from SAP-1005, the model reaches 6.99% CER and 10.71% WER on the dev-unshared subset, and its cross-etiology evaluation on TORGO yields 25.08% CER and 39.56% WER. The authors interpret the TORGO result as evidence that the model captures shared dysarthric speech characteristics rather than PD-specific artifacts. They also find that error rates rise with severity, from 4.73% CER for very low severity to 20.71% CER for high severity, and that spontaneous speech is much harder than digital-assistant commands, with WER of 19.29% versus 7.92%.

Load-bearing premise

The load-bearing assumption is that the SAP-1005 train and dev-unshared sets share no speakers and no text prompts, so that the reported error rates measure generalization to truly unseen speakers; if the official split leaks speakers or prompts, the headline numbers are optimistic.

Editorial extensions

If this is right

  • A single fine-tuned Whisper model can serve as a strong speaker-independent baseline for dysarthric ASR without per-speaker adaptation.
  • Learned representations transfer across dysarthria types: a PD-trained model partially handles CP and ALS speech, suggesting shared acoustic regularities across etiologies.
  • Performance degrades sharply with severity and with spontaneous speech, identifying the high-severity and spontaneous-speech regimes as the main remaining challenges.
  • Long spontaneous utterances remain a failure mode; chunking mitigates but does not eliminate Whisper's 30-second receptive-field hallucinations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the speaker and prompt isolation in SAP-1005 holds, then Whisper's pretraining already supplies enough phonetic generality that a small amount of fine-tuning on one dysarthria population transfers to others; a testable extension is to add a small amount of CP/ALS data and measure whether cross-etiology WER drops proportionally.
  • The reported comparison to the earlier wav2vec 2.0 baseline is not apples-to-apples because the earlier system was evaluated on a different portion of the corpus; a matched evaluation on the same dev-unshared partition would settle which backbone is actually better.
  • Because the official SAP-1005 test set is reserved for a competition, the dev-unshared numbers may be optimistic relative to a true held-out test, and readers should treat them as indicative until official test-set results appear.
  • Some high-severity spontaneous utterances show error rates exceeding 100%, meaning the model sometimes produces transcripts longer than the reference; reporting CER alone may hide such hallucination behavior.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper presents a speaker-independent dysarthric speech recognition system based on fine-tuning the multilingual Whisper medium model on the SAP-1005 dataset of Parkinson's disease speech. The authors use the dev unshared subset as a test set and the dev shared subset as validation, reporting a CER of 6.99% and WER of 10.71% on dev unshared, with breakdowns by severity and sentence category. They also evaluate the SAP-fine-tuned model on continuous speech from the TORGO dataset (cerebral palsy and ALS) without further adaptation, reporting an average CER of 25.08% and WER of 39.56%, and interpret this as evidence of cross-etiology generalization.

Significance. If the speaker-independence claim holds, the Whisper-based system provides a strong and useful baseline for the recently released SAP-1005 dataset and a benchmark for cross-etiology transfer from Parkinson's disease to CP/ALS dysarthria. The stratified results by severity and sentence category are informative, and the cross-etiology TORGO results are a valuable data point for the community. However, the validity of the headline numbers depends on the exact speaker partition between training, validation, and test, which is not established in the manuscript; the paper also relies on an admitted comparison to a different evaluation split.

major comments (4)
  1. [Section IV-B and Section III-A] The speaker-independent claim is not fully established because the validation and test sets may share speakers. Section III-A defines 'shared' and 'unshared' partitions by text-prompt overlap with the training set, not by speaker identity, and Section IV-B states that the dev shared set is used as the validation set for checkpoint selection while the dev unshared set is used as the test set. Since the 21-speaker dev set is split into shared and unshared portions, the same speakers can appear in both partitions. If so, selecting the best checkpoint on dev shared tunes the model to those speakers, and the reported 6.99% CER / 10.71% WER on dev unshared would overstate speaker-independent performance. The authors should verify speaker-ID disjointness between dev shared and dev unshared, and ideally provide per-speaker results or a validation split that is speaker-disjoint from the test speakers.
  2. [Section V-A and Table I] The headline comparison with Zheng et al.'s 26.92% WER is not apples-to-apples, as the manuscript itself admits in the text: 'the comparison is limited by the fact that we evaluated different portions of the dataset.' Reporting a 60.24% relative improvement over a method evaluated on a different split overstates the contribution. The authors should either reproduce the baseline on the exact same split (e.g., dev shared or dev unshared with comparable settings) or clearly label the comparison as indicative rather than as a relative improvement.
  3. [Tables I and II] The reported aggregate error rates lack confidence intervals, per-speaker breakdowns, and significance tests. In Table I, the overall averages (6.99% CER, 10.71% WER) are computed across severity groups that themselves vary widely (e.g., High severity CER 20.71% vs. Very Low 4.73%), so the average is sensitive to the composition of the dev unshared set. In Table II, per-speaker WER on TORGO ranges from 2.12% (M03) to 76.4% (M04), so the overall average of 39.56% is highly variable across only eight speakers. At minimum, per-speaker results and bootstrap confidence intervals are needed to assess the stability of the speaker-independent and cross-etiology claims.
  4. [Section IV-B] The chunking procedure for utterances longer than 30 seconds is described as splitting into 30-second chunks with a 5-second overlap, with each chunk decoded individually and the transcripts concatenated. The paper does not specify how the overlapped regions are resolved (e.g., deduplication of repeated words at boundaries) and does not provide a comparison against a no-chunking baseline or against different chunking overlaps. Since the authors attribute residual errors in spontaneous speech partly to hallucination in long utterances, an ablation would help establish the contribution of the chunking step.
minor comments (5)
  1. [Section III-A] The sentence 'these prompts are recorded by entirely different speakers, making SAP-1005 a speaker-independent dataset' conflates prompt sharing with speaker independence; the speaker-independent property requires explicit speaker disjointness between the train set and the dev/test sets, which is not described.
  2. [Figure 2] The legend order 'HMLVL' on the y-axis is not self-explanatory; the authors should spell out the severity levels or provide a note that H, M, L, VL denote High, Median (or Moderate), Low, and Very Low.
  3. [Section V-A] The text states that 'error rates exceeding 100% highlight these challenges' in relation to Figure 1; the figure's y-axis extends beyond 100%, but the caption should explicitly mention that CER and WER can exceed 100% due to insertions.
  4. [References] Reference [28] is listed as 'accepted' without a venue or page numbers, and reference [25] lacks university and year details; these should be completed for reproducibility.
  5. [Section V-B] The TORGO section reports averages over severity groups but does not state how many utterances per speaker were used in the evaluation; this information is needed to interpret the per-speaker error rates and the overall average.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: SAP-1005 and TORGO results are honest held-out measurements; only a minor model-choice self-citation is present.

full rationale

The central numbers (SAP-1005 CER 6.99%/WER 10.71% and TORGO CER 25.08%/WER 39.56%) are test-set measurements, not quantities derived by construction from a fitted parameter. The model is fine-tuned on the SAP-1005 train split and the validation checkpoint is selected on the dev-shared set; the reported results are on the dev-unshared set (Sections IV-B and V-A), and TORGO is an external dataset not used during training or checkpoint selection (Sections III-B, V-B). No equation in the paper defines a predicted error rate as a function of the fitted weights; the self-attention and beam-search equations are standard model descriptions, not a derivation of the reported scores. The only self-citation is [28], used to justify choosing the medium multilingual Whisper model (Section IV-B). That citation is a model-selection heuristic from the authors' prior benchmark, not a load-bearing premise asserting the specific SAP-1005/TORGO results, and the reported values would remain valid measurements even if the model had been chosen for a different reason. The potential concern that dev-shared (validation) and dev-unshared (test) may contain the same PD speakers is a data-split validity caveat, not a circularity: the paper does not exhibit the same-speaker overlap, and even if present, it would be evaluation leakage rather than the target quantity being equivalent to an input by construction. Accordingly, no circular step is identified; score 2 reflects only the minor non-load-bearing self-citation.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

No theoretical derivation is attempted; the evaluation rests on standard ML assumptions about data splits, metric comparability, preprocessing choices, and severity-label validity. The free parameters are the model variant, decoding settings, learning rate, and chunking strategy, none of which are ablated.

free parameters (4)
  • Whisper model variant and size = medium multilingual
    Chosen based on the authors' previous comparison on TORGO/UASpeech, not on SAP-1005; the central results depend on this choice.
  • Decoding hyperparameters = num_beams=10, no_repeat_ngram_size=3, length_penalty=1.0
    Set by hand in Section IV-B; no ablation is reported, so the headline numbers are conditional on these values.
  • Fine-tuning learning rate = 1e-5
    Reported in Section IV-B; number of epochs and batch size are not given, so the effective regularization is under-specified.
  • Chunking overlap = 30-second chunks with 5-second overlap
    Introduced to handle utterances longer than 30 seconds; affects long spontaneous speech results.
assumptions (4)
  • domain assumption The official SAP-1005 train/dev split is speaker-disjoint, and the dev-unshared test set contains no prompts seen in training.
    Invoked in Sections III-A and IV-B to justify the speaker-independent claim; the paper relies on dataset documentation rather than verifying speaker IDs.
  • domain assumption CER/WER are computed with a standard, comparable normalization of punctuation and casing.
    No normalization procedure is described in Sections IV-B or V, yet the numbers are compared with Zheng et al.'s published WER.
  • ad hoc to paper Whisper's 30-second chunking with 5-second overlap and simple concatenation preserves the content of long utterances.
    Section IV-B introduces this preprocessing; if it splits words or loses context, long-utterance results (especially spontaneous speech) would be affected.
  • domain assumption Severity labels derived from a wav2vec2 model's CER are a valid proxy for dysarthria severity.
    Section III-A defines VL/L/M/H from the CER of a LibriSpeech-finetuned wav2vec2 model; the severity-stratified analysis in Section V-A depends on these labels.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Cross-Etiology and Speaker-Independent Dysarthric Speech Recognition." pith.science (2026). https://pith.science/paper/U7GY7RI4

@misc{pith2026250114994,
  author       = {Pith},
  title        = {Pith review of: Robust Cross-Etiology and Speaker-Independent Dysarthric Speech Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U7GY7RI4}},
  note         = {Machine review of arXiv:2501.14994}
}
read the original abstract

In this paper, we present a speaker-independent dysarthric speech recognition system, with a focus on evaluating the recently released Speech Accessibility Project (SAP-1005) dataset, which includes speech data from individuals with Parkinson's disease (PD). Despite the growing body of research in dysarthric speech recognition, many existing systems are speaker-dependent and adaptive, limiting their generalizability across different speakers and etiologies. Our primary objective is to develop a robust speaker-independent model capable of accurately recognizing dysarthric speech, irrespective of the speaker. Additionally, as a secondary objective, we aim to test the cross-etiology performance of our model by evaluating it on the TORGO dataset, which contains speech samples from individuals with cerebral palsy (CP) and amyotrophic lateral sclerosis (ALS). By leveraging the Whisper model, our speaker-independent system achieved a CER of 6.99% and a WER of 10.71% on the SAP-1005 dataset. Further, in cross-etiology settings, we achieved a CER of 25.08% and a WER of 39.56% on the TORGO dataset. These results highlight the potential of our approach to generalize across unseen speakers and different etiologies of dysarthria.

Figures

Figures reproduced from arXiv: 2501.14994 by the authors.

Figure 1
Figure 1. Strip plot showing error rate of each utterance in dev [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Experimental results across different severity levels and sentence [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. A Self-Training Approach for Whisper to Enhance Long Dysarthric Speech Recognition

    cs.SD 2025-06 conditional novelty 6.0 of 10

    An iterative segmentation-based self-training method for Whisper improved long dysarthric speech recognition and achieved second place in both WER and SemScore at the SAP Challenge.

Reference graph

Works this paper leans on

41 extracted references · 38 canonical work pages · cited by 1 Pith paper

  1. [1]

    On the use of Hidden Markov Modelling for recognition of dysarthric speech,

    J. Deller Jr, D. Hsu, and L. J. Ferrier, “On the use of Hidden Markov Modelling for recognition of dysarthric speech,” Computer Methods and Programs in Biomedicine , vol. 35, no. 2, pp. 125–139, 1991

    work page 1991

  2. [2]

    Study of the Performance of Automatic Speech Recognition Systems in Speakers with Parkinson’s Disease,

    L. Moro-Velazquez, J. Cho, S. Watanabe, M. A. Hasegawa-Johnson, O. Scharenborg, H. Kim, and N. Dehak, “Study of the Performance of Automatic Speech Recognition Systems in Speakers with Parkinson’s Disease,” in Interspeech, vol. 9, 2019, pp. 3875–3879

    work page 2019

  3. [3]

    Speech errors in Cantonese speaking adults with cerebral palsy,

    T. L. Whitehill and V . Ciocca, “Speech errors in Cantonese speaking adults with cerebral palsy,” Clinical linguistics & phonetics , vol. 14, no. 2, pp. 111–130, 2000

    work page 2000

  4. [4]

    Speech deterioration in amyotrophic lateral sclerosis (ALS) after man- ifestation of bulbar symptoms,

    T. Makkonen, H. Ruottinen, R. Puhto, M. Helminen, and J. Palmio, “Speech deterioration in amyotrophic lateral sclerosis (ALS) after man- ifestation of bulbar symptoms,” International journal of language & communication disorders, vol. 53, no. 2, pp. 385–392, 2018

    work page 2018

  5. [5]

    An investigation of different degrees of dysarthric speech as input to speaker-adaptive and speaker-dependent recognition systems,

    P. Raghavendra, E. Rosengren, and S. Hunnicutt, “An investigation of different degrees of dysarthric speech as input to speaker-adaptive and speaker-dependent recognition systems,” Augmentative and Alternative Communication, vol. 17, no. 4, pp. 265–275, 2001

    work page 2001

  6. [6]

    Optimising speaker-dependent feature extraction parameters to improve automatic speech recognition performance for people with dysarthria,

    M. Marini, N. Vanello, and L. Fanucci, “Optimising speaker-dependent feature extraction parameters to improve automatic speech recognition performance for people with dysarthria,” Sensors, vol. 21, no. 19, p. 6460, 2021

    work page 2021

  7. [7]

    Two-step acoustic model adaptation for dysarthric speech recognition,

    R. Takashima, T. Takiguchi, and Y . Ariki, “Two-step acoustic model adaptation for dysarthric speech recognition,” in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2020, pp. 6104–6108

    work page 2020

  8. [8]

    Comparing speaker-dependent and speaker-adaptive acous- tic models for recognizing dysarthric speech,

    F. Rudzicz, “Comparing speaker-dependent and speaker-adaptive acous- tic models for recognizing dysarthric speech,” in Proceedings of the 9th international ACM SIGACCESS conference on Computers and accessibility, 2007, pp. 255–256

    work page 2007

Show all 41 references
  1. [9]

    Dysarthric speech transformer: A sequence-to-sequence Dysarthric speech recognition system,

    S. R. Shahamiri, V . Lal, and D. Shah, “Dysarthric speech transformer: A sequence-to-sequence Dysarthric speech recognition system,” IEEE Transactions on Neural Systems and Rehabilitation Engineering , vol. 31, pp. 3407–3416, 2023

    2023

  2. [10]

    Enhancing dysarthric speech recognition for unseen speakers via prototype-based adaptation,

    S. Wang, S. Zhao, J. Zhou, A. Kong, and Y . Qin, “Enhancing dysarthric speech recognition for unseen speakers via prototype-based adaptation,” in Proc. Interspeech, 2024, pp. 1741–1744

    2024

  3. [11]

    wav2vec 2.0: A framework for self-supervised learning of speech representations,

    A. Baevski, Y . Zhou, A. Mohamed, and M. Auli, “wav2vec 2.0: A framework for self-supervised learning of speech representations,” Advances in neural information processing systems , vol. 33, pp. 12 449– 12 460, 2020

    2020

  4. [12]

    wav2vec: Unsupervised pre-training for speech recognition,

    S. Schneider, A. Baevski, R. Collobert, and M. Auli, “wav2vec: Unsupervised pre-training for speech recognition,” arXiv preprint arXiv:1904.05862, 2019

    1904 arXiv

  5. [13]

    Robust speech recognition via large-scale weak supervi- sion,

    A. Radford, J. W. Kim, T. Xu, G. Brockman, C. McLeavey, and I. Sutskever, “Robust speech recognition via large-scale weak supervi- sion,” in International conference on machine learning (ICML) . PMLR, 2023, pp. 28 492–28 518

    2023

  6. [14]

    Improved meta learning for low resource speech recognition,

    S. Singh, R. Wang, and F. Hou, “Improved meta learning for low resource speech recognition,” in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2022, pp. 4798–4802

    2022

  7. [15]

    A Novel Self-training Approach for Low-resource Speech Recognition,

    S. Singh, F. Hou, and R. Wang, “A Novel Self-training Approach for Low-resource Speech Recognition,” in Proc. Interspeech, 2023, pp. 1588–1592

    2023

  8. [16]

    Speech Vision: An end-to-end deep learning-based dysarthric automatic speech recognition system,

    S. R. Shahamiri, “Speech Vision: An end-to-end deep learning-based dysarthric automatic speech recognition system,” IEEE Transactions on Neural Systems and Rehabilitation Engineering , vol. 29, pp. 852–861, 2021

    2021

  9. [17]

    Prosody-based measures for automatic severity assessment of dysarthric speech,

    A. Hernandez, S. Kim, and M. Chung, “Prosody-based measures for automatic severity assessment of dysarthric speech,” Applied Sciences , vol. 10, no. 19, p. 6999, 2020

    2020

  10. [18]

    Data augmentation using healthy speech for dysarthric speech recognition

    B. Vachhani, C. Bhat, and S. K. Kopparapu, “Data augmentation using healthy speech for dysarthric speech recognition.” in proc. Interspeech, 2018, pp. 471–475

    2018

  11. [19]

    Data augmenta- tion techniques for transfer learning-based continuous dysarthric speech recognition,

    T. Mariya Celin, P. Vijayalakshmi, and T. Nagarajan, “Data augmenta- tion techniques for transfer learning-based continuous dysarthric speech recognition,” Circuits, Systems, and Signal Processing , vol. 42, no. 1, pp. 601–622, 2023

    2023

  12. [20]

    Simulating dysarthric speech for training data augmentation in clinical speech applications,

    Y . Jiao, M. Tu, V . Berisha, and J. Liss, “Simulating dysarthric speech for training data augmentation in clinical speech applications,” in IEEE international conference on acoustics, speech and signal processing (ICASSP). IEEE, 2018, pp. 6009–6013

    2018

  13. [21]

    Phonetic analysis of dysarthric speech tempo and applications to robust personalised dysarthric speech recognition,

    F. Xiong, J. Barker, and H. Christensen, “Phonetic analysis of dysarthric speech tempo and applications to robust personalised dysarthric speech recognition,” in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2019, pp. 5836–5840

    2019

  14. [22]

    Speaker adaptation for Wav2vec2 based dysarthric ASR,

    M. K. Baskar, T. Herzig, D. Nguyen, M. Diez, T. Polzehl, L. Burget, J. ˇCernock`y et al. , “Speaker adaptation for Wav2vec2 based dysarthric ASR,” in Proc. Interspeech 2022 , 2022, pp. 3403–3407

    2022

  15. [23]

    Cross-lingual self-supervised speech rep- resentations for improved dysarthric speech recognition,

    A. Hernandez, P. A. P ´erez-Toro, E. N ¨oth, J. R. Orozco-Arroyave, A. Maier, and S. H. Yang, “Cross-lingual self-supervised speech rep- resentations for improved dysarthric speech recognition,” arXiv preprint arXiv:2204.01670, 2022

    2022 arXiv

  16. [24]

    Exploring the Impact of Fine-Tuning the Wav2vec2 Model in Database-Independent Detection of Dysarthric Speech,

    F. Javanmardi, S. R. Kadiri, and P. Alku, “Exploring the Impact of Fine-Tuning the Wav2vec2 Model in Database-Independent Detection of Dysarthric Speech,” IEEE Journal of Biomedical and Health Infor- matics, vol. 28, no. 8, 2024

    2024

  17. [25]

    Transfer Learning Using L2 Speech to Improve Au- tomatic Speech Recognition of Dysarthric Speech,

    H. Steinmetz, “Transfer Learning Using L2 Speech to Improve Au- tomatic Speech Recognition of Dysarthric Speech,” Master’s thesis, University of Washington, 2023

    2023

  18. [26]

    Noise Robust Whisper Features for Dysarthric Severity-Level Classification,

    S. Rathod, M. Charola, and H. A. Patil, “Noise Robust Whisper Features for Dysarthric Severity-Level Classification,” in International Conference on Pattern Recognition and Machine Intelligence . Springer, 2023, pp. 708–715

    2023

  19. [27]

    Transfer Learning Using Whisper for Dysarthric Automatic Speech Recognition,

    ——, “Transfer Learning Using Whisper for Dysarthric Automatic Speech Recognition,” in International Conference on Speech and Com- puter. Springer, 2023, pp. 579–589

    2023

  20. [28]

    A Comprehensive Performance Evaluation of Whisper Models in Dysarthric Speech Recognition,

    S. Singh, Z. Zhong, Q. Wang, C. Mendes, M. Hasegawa-Johnson, W. Abdulla, and S. R. Shahamiri, “A Comprehensive Performance Evaluation of Whisper Models in Dysarthric Speech Recognition,” in Proceedings of the International Conference on Neural Information Processing (ICONIP), ...

    2024

  21. [29]

    The TORGO database of acoustic and articulatory speech from speakers with dysarthria,

    F. Rudzicz, A. K. Namasivayam, and T. Wolff, “The TORGO database of acoustic and articulatory speech from speakers with dysarthria,” Language resources and evaluation , vol. 46, pp. 523–541, 2012

    2012

  22. [30]

    Dysarthric speech database for universal access research,

    H. Kim, M. Hasegawa-Johnson, A. Perlman, J. R. Gunderson, T. S. Huang, K. L. Watkin, and S. Frame, “Dysarthric speech database for universal access research,” in Proc. Interspeech, 2008, pp. 1741–1744

    2008

  23. [31]

    UIUC Leads Effort to Make Speech Recognition Technology More Inclusive,

    L. O’Malley, “UIUC Leads Effort to Make Speech Recognition Technology More Inclusive,” Insight Into Diversity, November 6 2023, accessed: 2024-09-04. [Online]. Available: https://www.insightintodiversity.com/ uiuc-leads-effort-to-make-speech-recognition-technology-more-inclusive/

    2023

  24. [32]

    Fine-Tuning Auto- matic Speech Recognition for People with Parkinson’s: An Effective Strategy for Enhancing Speech Technology Accessibility,

    X. Zheng, B. Phukon, and M. Hasegawa-Johnson, “Fine-Tuning Auto- matic Speech Recognition for People with Parkinson’s: An Effective Strategy for Enhancing Speech Technology Accessibility,” in Proc. Interspeech, 2024, pp. 2485–2489

    2024

  25. [33]

    Articulatory and bottleneck features for speaker-independent ASR of dysarthric speech,

    E. Yılmaz, V . Mitra, G. Sivaraman, and H. Franco, “Articulatory and bottleneck features for speaker-independent ASR of dysarthric speech,” Computer Speech & Language , vol. 58, pp. 319–334, 2019

    2019

  26. [34]

    Improved ASR Performance for Dysarthric Speech Using Two-stage Data Augmentation,

    C. Bhat, A. Panda, and H. Strik, “Improved ASR Performance for Dysarthric Speech Using Two-stage Data Augmentation,” in proc. in- terspeech, 2022, pp. 46–50

    2022

  27. [35]

    SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition,

    D. S. Park, W. Chan, Y . Zhang, C.-C. Chiu, B. Zoph, E. D. Cubuk, and Q. V . Le, “SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition,” Proc. Interspeech, p. 2613, 2019

    2019

  28. [36]

    Sequence transduction with recurrent neural networks,

    A. Graves, “Sequence transduction with recurrent neural networks,” arXiv preprint arXiv:1211.3711 , 2012

    2012 arXiv

  29. [37]

    The effect of speech and speech intelligibility on task performance,

    N. Venetjoki, A. Kaarlela-Tuomaala, E. Keskinen, and V . Hongisto, “The effect of speech and speech intelligibility on task performance,” Ergonomics, vol. 49, no. 11, pp. 1068–1091, 2006

    2006

  30. [38]

    Clinical and acoustical variability in hypokinetic dysarthria,

    E. J. Metter and W. R. Hanson, “Clinical and acoustical variability in hypokinetic dysarthria,” Journal of communication disorders , vol. 19, no. 5, pp. 347–366, 1986

    1986

  31. [39]

    Spastic dysarthria in cerebral palsy,

    K. C. Hustad and K. Sassano, “Spastic dysarthria in cerebral palsy,” Journal of Medical, Speech-Language Pathology , vol. 10, no. 4, pp. 287–292, 2002

    2002

  32. [40]

    Disorders of communication: dysarthria,

    P. Enderby, “Disorders of communication: dysarthria,” Handbook of clinical neurology, vol. 110, pp. 273–281, 2013

    2013

  33. [41]

    Ataxic dysarthria,

    R. D. Kent, J. F. Kent, J. R. Duffy, J. E. Thomas, G. Weismer, and S. Stuntebeck, “Ataxic dysarthria,” Journal of speech, language, and hearing research, vol. 43, no. 5, pp. 1275–1289, 2000

    2000

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.